Early tumor detection is a critically important goal in oncologic imaging because it would enable treatment (or redirection of treatment) at earlier stages of disease. However, the challenge is imaging small collections of tumor cells because conventional in vivo imaging methods suffer from limited resolution and tumor contrast. MRI, which presents exquisite anatomical detail and many contrast mechanisms, still often proves incapable of detecting small clusters of tumor cells. Our strategy will be to develop a magnetic resonance imaging (MRI) agent with great amplification potential and to use magnetic resonance (MR) acquisition and image processing techniques to provide contrast sufficient to detect even very small tumors. Our proposal builds upon substantial preliminary data and brings together significant multi-disciplinary expertise in probe chemistry, tumor models, tumor biology, unique 3D microscopic cryo-imaging, advanced MR techniques, and quantitative image analysis. To date, we have discovered a novel molecular imaging strategy by targeting extracellular fragments of PTP? abundantly found in the tumor microenvironment; created mouse orthotopic models of gliomas; developed cryo-imaging methods to visualize and quantify tumor size, cell dispersal, white matter tracts and blood vessel density in 3D brain reconstructions; and discovered that a fluorescent PTP? probe quickly labeled main tumor as well as dispersed cells and even single migrating cells up to 3.5 mm away from the main tumor mass. Recently preliminary studies using a gadolinium conjugated PTP? probe indicate that we can see small tumors using MRI. We seek funding to demonstrate delineation of the dispersing tumor boundary and detection of tiny clusters of cancer cells using the PTP? molecular imaging probe by MRI. Even a skilled radiologist will not have the confidence to specifically diagnose a tumor until it is bigger than 5x5x5 mm3 even though typical spatial resolution is about 1x1x5 mm3 because of the non-specific, non-quantitative nature of the MR signal. Our approach will be to greatly increase contrast of the MR signal relative to background anatomical variations through amplification characteristics of PTP? MRI probe (PTP?-Gd), hardware, acquisition, and software improvements. We hypothesize that we can specifically identify and characterize tumors 2-3 orders of magnitude smaller than currently possible through the use of the PTP? molecular targeting agent combined with high resolution and quantitative MR. The Specific Aims are: 1. Optimize and test the ability of the molecular PTP?-Gd probe to detect brain tumors in orthotopic xenografts and compare to conventional brain tumor MRI. 2. Determine the ability of PTP?-Gd to accurately image dispersing brain tumors as compared to gold-standard GFP-labeled tumor from microscopic cryo-imaging. 3. Optimize quantitative MRI using a clinically feasible (3T) dual-agent approach to test the limits for detecting small isolated brain tumors in a very highly dispersing tumor model.